Document of The World Bank

FOR OFFICIAL USE ONLY

Report No: PAD1978

INTERNATIONAL DEVELOPMENT ASSOCIATION

PROJECT PAPER

ON A

PROPOSED ADDITIONAL IDA GRANT IN THE AMOUNT OF SDR 10.8 MILLION

(US$15 MILLION EQUIVALENT)

AND A

PROPOSED ADDITIONAL IDA GRANT IN THE AMOUNT OF SDR 14.3 MILLION

(US$20 MILLION EQUIVALENT) IN CRISIS RESPONSE WINDOW RESOURCES

TO THE

REPUBLIC OF MADAGASCAR

FOR A

SOCIAL SAFETY NET PROJECT

October 28, 2016

Social Protection and Labor Global Practice Africa Region

This document has a restricted distribution and may be used by recipients only in the performance of their official duties. Its contents may not otherwise be disclosed without World Bank authorization.

Pub

lic D

iscl

osur

e A

utho

rized

Pub

lic D

iscl

osur

e A

utho

rized

Pub

lic D

iscl

osur

e A

utho

rized

Pub

lic D

iscl

osur

e A

utho

rized

CURRENCY EQUIVALENTS (Exchange Rate Effective September 30, 2016)

Currency Unit = Malagasy Ariary (MGA)

US$1 = MGA 3111 US$ 1 = SDR 0.71642989

FISCAL YEAR

January 1 – December 31

ABBREVIATIONS AND ACRONYMS ACN Community Nutrition Agents (Agents Communautaire de Nutrition) AfDB African Development Bank AF Additional Financing BNGRC National Office for Risk and Disaster Management (Bureau National

de Gestion des Risques et des Catastrophes) CCT Conditional Cash Transfer FAO Food and Agriculture Organization EPM Periodic Household Survey (Enquête Périodique auprès des

Ménages) ESMF Environmental and Social Management Framework FID Intervention Fund for Development (Fonds d’Intervention pour le

Développement) FM Financial Management GRS Grievance Redress Service HDCT Human Development Cash Transfer IDA International Development Association INSTAT National Statistics Office (Institut National de la statistique de

Madagascar)

MAM Moderate Acute Malnutrition MEN Ministry of Education (Ministère de l'Education Nationale) MoH Ministry of Health MPSPPW Ministry of Population, Social Protection and Promotion of Women NGOs Nongovernmental Organizations ONN National Office for Nutrition (Office National de Nutrition) PAUSENS Madagascar Emergency Support to Critical Education, Health and

Nutrition Services Project (Projet d’Appui d’Urgence aux Services Essentiels de l’Education, de la Santé, et de la Nutrition)

PDO Project Development Objective PNNC National Community Nutrition Program (Programme National de la

Nutrition Communautaire) SAM Severe Acute Malnutrition SSNP Social Safety Net Project UN UN OCHA

United Nations UN Office for the Coordination of Humanitarian Affairs

UNICEF United Nations Children's Fund USAID United States Agency for International Development U-PNNC Project Implementation Unit for the National Community Nutrition

Program (Unité Programme National de Nutrition Communautaire) WFP World Food Programme WHO World Health Organization

Regional Vice President: Makhtar Diop Country Director: Mark R. Lundell

Senior Global Practice Director: Michal Rutkowski Country Manager: Coralie Gevers

Practice Manager/Manager: Dena Ringold Task Team Leaders: Andrea Vermehren/Jumana Qamruddin

MADAGASCAR

AF Social Safety Net (P160554)

CONTENTS I. Introduction ............................................................................................................................ 4

II. Background and Rationale ..................................................................................................... 4

III. Proposed Changes ................................................................................................................ 16

IV. Appraisal Summary ............................................................................................................. 21

V. World Bank Grievance Redress ........................................................................................... 25

Annex I: Revised Results Framework and Monitoring ............................................................. 27

Annex II: Detailed Description of Modified or New Project Activities ..................................... 33

Annex III: Geographic Targeting of the Additional Targeting .................................................... 40

Annex IV: Revised Estimate of Project Costs (US$, millions) ................................................... 42

Annex V: Summary of the Human Development Cash Transfer Program (HDCT) .................. 43

Annex VI: Summary of the National Community Nutrition Program (PNNC) ........................... 45

Annex VII: Key Results from the Health and Education Public Expenditure Review ................ 47

i

ADDITIONAL FINANCING DATA SHEET Madagascar

AF Social Safety Net ( P160554 ) AFRICA GSP01

.

Basic Information – Parent Parent Project ID: P149323 Original EA Category: B - Partial Assessment

Current Closing Date: 30-Sep-2020

Basic Information – Additional Financing (AF)

Project ID: P160554 Additional Financing Type (from AUS): Scale Up

Regional Vice President: Makhtar Diop Proposed EA Category:

Country Director: Mark R. Lundell Expected Effectiveness Date: 30-Dec-2016

Senior Global Practice Director: Michal J. Rutkowski Expected Closing Date: 30-Sep-2020

Practice Manager/Manager: Dena Ringold Report No: PAD1978

Team Leader(s): Andrea Vermehren, Jumana N. Qamruddin

Borrower Organization Name Contact Title Telephone Email Ministry of Finance and Budget (MFB)

Alexandre Randrianasolo

Secretaire General 2610340115243

Project Financing Data - Parent ( Social Safety Net Project-P149323 ) (in USD Million) Key Dates

Project Ln/Cr/TF Status Approval Date Signing Date Effectiveness

Date Original Closing Date

Revised Closing Date

P149323 IDA-57080 Effective 16-Sep-2015 13-Nov-2015 10-Feb-2016 30-Sep-2020 30-Sep-2020

Disbursements

Project Ln/Cr/TF Status Currency Original Revised Cancelled Disbursed Undisbursed % Disbursed

P149323 IDA-57080 Effective XDR 28.50 28.50 0.00 5.74 22.76 20.15

ii

Project Financing Data - Additional Financing AF Social Safety Net (P160554) (in USD Million) [ ] Loan [ ] Grant [X] IDA Grant [ ] Credit [ ] Guarantee [ ] Other Total Project Cost: 35.00 Total Bank Financing: 35.00 Financing Gap: 0.00

Financing Source – Additional Financing (AF) Amount IDA Grant 15.00 IDA Grant from CRW 20.00 Financing Gap 0.00

Total 35.00

Policy Waivers Does the project depart from the CAS in content or in other significant respects? No

Explanation Does the project require any policy waiver(s)? No Explanation

Team Composition

Bank Staff

Name Role Title Specialization Unit Andrea Vermehren Team Leader

(ADM Responsible)

Lead Social Protection Specialist

GSP01

Jumana N. Qamruddin Team Leader Senior Health Specialist

Co TTL GHN01

Lova Niaina Ravaoarimino

Procurement Specialist (ADM Responsible)

Senior Procurement Specialist

GGO07

Enagnon Ernest Eric Adda

Financial Management Specialist

Sr Financial Management Specialist

GGO31

Abu Yadetta Hateu Peer Reviewer Sr Social Protection Specialist

GSP01

Irene Muthoni Ngatia Team Member Young Professional Economist AFRCE

Julia Rachel Ravelosoa

Team Member Senior Social Protection Economist

GSP01

Khurshid Banu Noorwalla

Team Member Program Assistant GSP07

iii

Menno Mulder-Sibanda

Peer Reviewer Sr Nutrition Spec. GHN07

Paul-Jean Feno Safeguards Specialist

Senior Environmental Specialist

GEN07

Peter F. B. A. Lafere Safeguards Specialist

Senior Social Development Specialist

GSU01

Siobhan McInerney-Lankford

Counsel Senior Counsel LEGAM

Ugo Gentilini Peer Reviewer Sr Social Protection Specialist

GSPGL

Valerie Rambeloson Team Member Consultant GHN01 Voahirana Hanitriniala Rajoela

Team Member Senior Health Specialist

Sr. HNP Specialist GHN01

Extended Team Name Title Location

Christophe Ribes Ros Safety Nets Consultant Mallorca, Spain Locations Country First Administrative

Division Location Planned Actual Comments

Madagascar Republic of Madagascar

Institutional Data

Parent ( Social Safety Net Project-P149323 ) Practice Area (Lead) Social Protection & Labor

Contributing Practice Areas

Additional Financing AF Social Safety Net ( P160554 ) Practice Area (Lead) Social Protection & Labor

Contributing Practice Areas Health, Nutrition & Population

Consultants (Will be disclosed in the Monthly Operational Summary) Consultants will be required

4

I. Introduction

1. This Project Paper seeks the approval of the Executive Directors to provide an additional financing (AF) to the Republic of Madagascar for the Social Safety Net Project (P149323— the ‘Project’ or ‘SSNP’) in an amount of SDR25.1 million (US$35 million equivalent) of which SDR14.3 million (US$20 million equivalent) is from the IDA Crisis Response Window1(CRW) and SDR10.8 million (US$15 million equivalent) is from the national IDA allocation. The proposed AF would scale up the existing SSNP through cash transfers and community nutrition services to address some of the urgent needs of the poorest population suffering from the severe drought in the South of Madagascar, exacerbated by El Niño. On September 26, 2016, the Government of Madagascar declared a humanitarian emergency for the region and requested support from the international community, including the World Bank, to provide a rapid and effective response for the most affected population.

2. No major changes to the project are proposed. The project development objective (PDO) of the original SSNP, which is “to support the Government in increasing the access of extremely poor households to safety net services and in laying the foundations for a social protection system”, will not change. The original key performance indicators will be maintained with some targets adjusted and indicators added at the PDO and intermediate outcome level to reflect the results expected from the proposed AF. The project components remain unchanged. Financial Management (FM) and procurement arrangements will be adjusted to reflect an additional implementing agency.

3. The proposed AF is being processed under condensed procedures provided for by OP 10.00, paragraph 12: Projects in Situation of Urgent Need of Assistance and Capacity Constraints, due to the drought emergency. This applies only to the use of the condensed procedures and does not include a deferral of safeguard requirements or the use of alternative fiduciary arrangements.

4. Financing for this AF meets all the criteria for CRW financing. It is triggered by an exceptional catastrophic event affecting a large number of extremely poor people. It is grounded in rigorous impact assessments undertaken by UN agencies and validated by the Government and Development Partners. It is aligned with the “Madagascar Deep South: Strategic Response Plan to the Extended Drought for 2016-2017” elaborated by the national emergency response agency and UN OCHA (UN Office for the Coordination of Humanitarian Affairs). Existing IDA resources have already been re-deployed but these are too limited to address the scale of the challenge.

II. Background and Rationale

5. Madagascar is one of the poorest countries in the world with respect to both the breadth and depth of poverty. Almost 78 percent (2015) of Madagascar’s population of 24.2 million people live on less than US$1.90 per day. These 19 million people make Madagascar the

1 Management informed the Executive Directors of its intention to allocate an indicative amount of USD20 million equivalent to support Madagascar’s response to the impact of the drought caused by El Niño at a technical briefing on October 12, 2016. See the note entitled "IDA Crisis Response Window Support for Lesotho, Madagascar, Malawi, and Mozambique for an El Niño Drought Response" IDA/SecM2016-0195, dated October 6, 2016 for additional information.

5

country with the sixth highest number of poor in the world.2 A stunningly high proportion of close to 60 percent of the population is extremely poor based on the minimum food intake methodology. This means that close to 13 million Malagasy people live on resources that fall below the cost of about 2,100 calories a day. Madagascar also has the world’s fourth highest rate of chronic malnutrition,3 with almost half of all children under five years stunted.

6. The ‘Deep South’ is the poorest and most isolated part of the country. Three regions belong to the ‘Grand Sud’: Androy, Anosy, and Atsimo Andrefana. An analysis of the socioeconomic situation4 of the eight drought-affected districts of the South confirms the severe underperformance of the region in all human development related aspects. Almost half of the children ages 6 to 10 years never go to school, alphabetization rates for people older than 15 years of age are low at 43.7 percent against the national average of 71.6 percent, and the average food consumption per capita is about one-third lower than the national average. Vaccination rates for children ages 12 to 23 months are between 31 percent and 37 percent, compared to a national average of 51 percent. The South is also characterized by a fragile social and political context (for example, trafficking of cattle/gold/gems and abuse of human rights) with increased insecurity due to roaming gangs and a general sense of lawlessness.

7. In addition, the South is one of the most underserved regions of the country with respect to public investment. The Public Expenditure Review of the health and education sectors conducted by the World Bank and the United Nations Children's Fund (UNICEF) in 2014/155 shows that the poorest regions have the lowest amount of spending on these key sectors (see annex VII). The low amount of education and health expenditure in rural areas, compared to urban areas, is indicative of large inequities in the distribution of public resources, mostly driven by the inequitable distribution of investments and human resources across the country. Cuts in public spending to social sectors during the crisis resulted in large increases in out-of-pocket household spending, particularly on education. This had far-reaching consequences for children’s enrollment in primary schools and learning outcomes, as well as the utilization of health services. Furthermore, investments in economic infrastructure (roads, water, electricity, etc.) have been largely absent in the last decades.

8. The more recent phenomenon of El Niño caused a prolonged, severe drought that particularly affected the already highly impoverished ‘Deep South’. The population of this region has already suffered through several successive years of poor crop yields, starting with a major locust invasion in 2013. The last two years have seen droughts that are more severe and broader in scope than predicted. Poor crop yields and the subsequent decimation of livestock have had a devastating impact on the population in the South. The Ministry of Agriculture in collaboration with the World Food Programme (WFP), the United Nations (UN) Food and Agriculture Organization (FAO), and UNICEF has been periodically assessing the situation over the past year. Data collected by the Government with support from WFP and UNICEF in February 2016 confirms the extraordinary situation:

2 This is after India, Nigeria, Democratic Republic of Congo, China, and Bangladesh. Grow, Invest, Insure: A game plan to end poverty by 2030, I. Gill, A. Revenga, Ch. Zeballos, July 2016. 3 The four highest rates are in Papua New Guinea, Burundi, Eritrea, and Madagascar. 4 Analyse de la situation socio-economique dans le sud de Madagascar, World Bank August 2016, unpublished draft. 5 Madagascar Public Expenditure Review for Health and Education, World Bank, 2015.

6

• Food insecurity. Approximately 80 percent of the population of the southern districts (1,140,000 people) is in a situation of food insecurity, of which more than half are severely food insecure. This is 14 percent higher than in March 2015. Crop yields will be nil for maize and at 20 percent of the historical average for cassava, the main staples of the poor in the South.

• Acute malnutrition. Nearly 45,000 children between 6 months and 5 years suffer from moderate acute malnutrition (MAM); 10,000 suffer from severe acute malnutrition (SAM) with faster rising trends in the past year. In a number of communes, MAM of children is now above the 15 percent mark signaling an emergency in nutritional outcomes. In some pockets of the Androy region, large food consumption gaps are likely, in line with emergency (Integrated Phase Classification Phase 4) expected outcomes between October 2016 and January 2017.

9. The poorest households have been particularly affected by the drought. In the 2010 Household Survey on Living Standards, the most common shocks reported by households were weather-related with drought being the most cited. The assets and income losses for those affected households amounted to an average of 80 percent of household consumption for the first quintile and as much as 150 percent for the second quintile.

Table 1: Top 10 Negative Shocks Reported by Households, 2010

2010 Shocks %

Drought 14.6 Flooding 10.5 Cyclone 9.7 Crop diseases 6.8 Late rains 6.4 Illness or loss of livestock 6.3 General rise in consumer prices 4.5 Illness of an adult household member 3.8 Increase in input prices 3.7 Other climatic or environmental shocks 3.6

Source. National Statistics Office (Institut National de la statistique de Madagascar, INSTAT), Periodic Household Survey (Enquête Périodique auprès des Ménages, EPM) 2005–2010.

7

Figure 1: Asset and Income Losses due to Shocks (2010)

Source: Madagascar Millennium Development Goals National Monitoring Survey 2012.

10. The United Nations Office for the Coordination of Humanitarian Affairs estimates that as much as US$110 million would be required over a period of 18 months to alleviate the dire situation in the South. This estimate is based on the rigorous national disaster assessment and evaluation undertaken by WFP, FAO, and UNICEF with the support of the Government and the members of the UN Humanitarian Country Team including the World Bank. Discussions with key ministries and results from WFP situation assessments indicate that the region’s population will need emergency support well into 2017 because of the high de-capitalization of the affected population over the past two-year drought. Thus an initial ‘Madagascar Deep South: Strategic Response Plan to the Extended Drought for 2016–2017’ has been developed by the National Office for Risk and Disaster Management (Bureau National de Gestion des Risques et des Catastrophes, BNGRC) and the United Nations Office for Coordination of Humanitarian Affairs, under the leadership of the Prime Minister’s Office and the Ministry of Interior. The BNGRC is also developing a medium-term recovery plan running until 2019 with the support of the Government and the development partners.

11. The Government is taking action within its limited fiscal means. Madagascar is coming out of a five-year crisis during which most international aid was interrupted and/or redirected toward non-governmental organizations (NGOs). The country has little fiscal space to respond to this emergency with its own funds: fiscal revenues are among the lowest in the world (10.4 percent of GDP) and foreign aid remains at a low level, even when compared to other countries with similar Country Policy and Institutional Assessment and development needs.6 The Government concluded a three-year Extended Credit Facility program with the IMF in July 2016 and will organize a donors roundtable in December 2016. Despite the tight fiscal situation, the Council of Government issued and adopted a ‘verbal notification’ from the BNGRC on the drought emergency in the South and established an ‘Emergency Commission’ under the leadership of the Prime Minister and the Ministry of Interior. They have taken the following actions:

• The BNGRC and the Ministry of Population, Social Protection, and Promotion of Women (MPSPPW) have led the coordination efforts across ministries and

6 See Madagascar: Systematic Country Diagnostic, 2015, chapter 3.

8

development partners in the capital and in the southern region where it recently opened a local office. It is mapping all humanitarian responses and is leading the effort to develop a short- and medium-term recovery strategy (see above).

• The BNGRC has organized the delivery of rice (provided by the Government) and other goods to the most affected communities.

• The National Office for Nutrition (Office National de Nutrition, ONN) has redirected some of its own funding toward the South to ensure that an increasing number of community nutrition centers are operational.

• The Ministry of Agriculture, the Ministry of Health (MoH), and the ONN actively undertake periodic evaluations with the guidance and funding from FAO, WFP, UNICEF, and World Health Organization (WHO).

12. A number of development partners, including the World Bank, have been working in the South over the past couple of years and are now allocating funds toward the drought emergency response. The European Union’s European Civil Protection and Humanitarian Aid Operations, France, Kuwait, Norway, Switzerland, UN Agencies (for example, through a contribution from the Central Emergency Response Fund), United States Agency for International Development (USAID), and African Development Bank have committed as much as US$20 million, with possibly another US$8 million to come. International NGOs have been able to bring in an additional US$10 million. However, this is insufficient in scope, geographical coverage, and population coverage.

13. Over the past year, the World Bank has used the Contingent Emergency Response Components under three existing projects and has reallocated close to US$3.5 million to respond to the emerging crisis:7 (a) the Emergency Global Partnership for Education Project redirected US$650,000 toward the school canteen program run by WFP; (b) the Madagascar Emergency Support to Critical Education, Health and Nutrition Services Project (PAUSENS, P131945) has reallocated US$1,200,000 to screen children under five years for acute malnutrition and supply food supplements for those children diagnosed as malnourished, and is allocating an additional US$1,500,000 to scale up food rations to targeted families in coordination with WFP; and (c) the Emergency Food Security and Social Protection Project (P147514) allocated US$100,000 for immediate cash-for-work activities in the affected region. These reallocations from the existing IDA portfolio are in addition to the US$13.5 million of the Emergency Food Security and Social Protection Project and the US$10 million in AF to the PAUSENS Project that was directed toward financing the three-year emergency response to the 2013 locust invasion.

14. Despite the combined efforts by the Government and development partners, the emergency response falls short of the estimated needs of the affected population. According to estimates by the World Bank, the main humanitarian agencies (USAID, WFP, and other UN agencies) cover about 250,000 people in the short term with food aid, cash-for-work activities, and so on (see table 2). They are in the process of raising funds to support about 300,000 people in the

7 In the spring of 2016, the Government expressed interest in putting in place an Immediate Response Mechanism. The operational manual is almost finalized and should be adopted before the end of 2016.

9

South in the short to medium term with mostly food and other early recovery support. Including the support from the proposed AF for the SSNP, still about one-third of the affected population would not receive any support. Therefore, the Government in collaboration with development partners, mostly from the UN agencies, continues its efforts to raise funds to close the remaining gap.

Table 2. Number of People Covered by the Main Emergency Response Programs in the Three Regions of the Deep South of Madagascar

Number of People Population in the South affected by the drought

1,027,663

USAID, WFP, FAO, and others 409,000 (June–December 2016)* USAID (through WFP, CRS, and others) 100,000 (October 2016–March 2017)* IDA AF SSNP (Crisis Response Window and IDA)

316,000 (November 2016–December 2019)

Coverage gap 218,663 (short and medium term) Source: Estimates by World Bank, various sources including USAID’s Famine Early Warning Systems Network June 2016–January 2017. Rationale for Additional Financing

15. The current situation of Madagascar’s South reflected in the increasing food insecurity and rising malnutrition rates of children, requires an effective and fast response. International evidence shows that the cost of inaction in such a situation is high. Moderate malnutrition increases the risk of mortality and its effects on children can be irreversible. If some children with MAM do not receive adequate support, they can progress toward SAM, severe wasting, and/or edema or severe stunting, both of which are life-threatening conditions. Stunting, in particular, is linked to negative long-term impacts on human development and economic outcomes, all of which disproportionately affect the poorest and most vulnerable population in the region. Given the dire situation, the abovementioned funding gap, and the long-term partnership and support to the country, the rationale for World Bank engagement is strong.

16. The AF to the SSNP is a rapid and efficient mechanism to respond to the emergency and support the recovery of the households. The SSNP already includes disaster response activities, provides regular cash transfers to extremely poor families, including in some of the affected areas, and has the capacity to quickly adapt and scale up the safety net interventions to confront the situation in the South. The envisaged activities of cash transfers combined with nutrition services fit under the objectives of the original operation. The AF is the fastest way to get much-needed income support and services directly to beneficiaries who have faced rapid erosion of their food security, health, and livelihoods.

17. Furthermore, over time, the AF provides an opportunity to connect the emergency response with a more resilience-oriented approach in the South as part of an adaptation/mitigation strategy. After the first year of urgent response through rapidly disbursing cash transfers, livelihood grants, and nutrition services focusing on treating MAM, the AF will convert the cash transfers into conditional cash transfers (CCTs) addressing, where possible, the

10

alarming health and education situation of children. In addition, the program implementing agencies will make an effort to coordinate the cash transfers with complementary services of other institutions, for example, the FAO, which is providing seeds, tools, and other inputs in the South just in time for the new planting season. Similarly, the AF will contribute to mitigate malnutrition with enhanced emergency nutrition services by including complementary activities related to nutrition, education, and malnutrition prevention. The mechanism for delivering nutrition services has proven to be resilient through the periods of cyclical fragility experienced in the country (for more information on Madagascar’s National Community Nutrition Program [Programme National de la Nutrition Communautaire, PNNC], see annex VI). This combined approach is expected to strengthen the households’ resilience and improve outcomes in the longer term.

18. The proposed AF will also leverage parallel financing from UNICEF. UNICEF has secured some emergency response funds of about US$1 million and is prepared to apply the same design as this AF, including cash transfers and nutrition services, specifically treatment of SAM,8 at a facility level. UNICEF is also co-financing the impact evaluation of the cash transfer program. The World Bank is also coordinating with the Government (BNGRC, MPSPPW, and so on), and with key development partners like USAID, FAO and WFP to ensure complementarity of resources with respect to geographic targeting and types of interventions.

Progress in Project Implementation

19. Launched in February 2016, the SSNP supports the Government in increasing the access of extremely poor households to safety net services and in laying the foundations for a social protection system (PDO). The US$40 million IDA credit was approved by the World Bank’s Board of Directors on September 16, 2015 with a project closing date of September 30, 2020. The project beneficiaries are extremely poor households living in communities in selected regions in the South and East of the country. The project’s disaster response subcomponent (Subcomponent 1.3) benefits communities nationwide based on the country’s system in determining disasters, coordinated by the BNGRC. Women particularly benefit from the SSNP, as they are the direct beneficiaries for most of the safety net activities.

20. The SSNP includes three components:

• Component 1: Building a Safety Net for the Poor in Selected Rural Areas (US$30.2 million)

o Subcomponent 1.1: Establishing a Productive Safety Net Program (US$14.7 million)

o Subcomponent 1.2: Expanding the Human Development Cash Transfer Pilot (US$11.5 million)

8 The management of SAM is through health care facilities with the support of UNICEF, and funding from the European Civil Protection and Humanitarian Aid Operations. The management of MAM is through the community nutrition sites, under the ONN/Unité Programme National de Nutrition Communautaire (U-PNNC) management.

11

o Subcomponent 1.3: Early Recovery Response to Natural Disasters (US$4 million)

• Component 2: Strengthening Safety Net Administration, Monitoring, and Social Accountability (US$6.5 million)

• Component 3: Building the Institutional Capacity for Coordination, Monitoring, and Evaluation of the Social Protection System (US$3.3 million)

21. Overall progress under the SSNP toward meeting the development objectives is Moderately Satisfactory. The project became effective in February 2016 and 20 percent of the credit proceeds have already been disbursed. All components have started implementation albeit a bit slower than planned because of several innovations introduced in the implementation of social safety net programs, and due to the fact that the MPSPPW is a new implementing agency and is building its capacity for coordination of the social protection sector. The drought in the South as well as other disasters (floods in the North and center) have demanded attention by the Ministry and the Intervention Fund for Development (Fonds d’Intervention pour le Développement, FID), both of which supported emergency response activities through cash-for-work in affected regions. The World Bank team is closely following up and supporting the implementation with technical assistance to ensure continued satisfactory progress.

Scope and Design of Additional Financing

22. The proposed AF will provide cash transfers, livelihood recovery grants, and nutrition services to the poorest households in the most drought-affected communities of the regions of Androy and Anosy of Madagascar. As depicted in figure 2, the AF will transition from emergency response to early recovery and mitigation through safety net and nutrition services.

Figure 2. From Emergency Response to Early Recovery

*Note: Plumpy’Sup is a nutritional supplement for malnourished children

Emergency Response

-Cash transfers and livelihood recovery grants -Emergency treatement of MAM through community nutrition sites including screening, Plumpy’Sup, protective food rations and household follow-up

Early Recovery-Conversion of cash transfers into CCTs -Livelihood recovery grants -Continuation of MAM treatment consisting of screening, Plumpy’Sup* and household follow-up-Enhanced focus on preventative nutrition measures for acute and chronic malnutrition/early stimulation at household level

Early Recovery

-CCTs -Focus on enhanced provision of preventative nutrition services/early stimulation activities

36 MONTHS

12

23. The AF will scale up the following components of the SSNP over a three-year period:

• Subcomponent 1.3: Early Recovery Response to Natural Disasters (original US$4 million, additional US$30.9 million).9 Scale-up through cash transfers, nutrition treatment and prevention services, and livelihood recovery grants. In light of the high rate of MAM among children under five years and the fact that the initial credit of the IDA-funded PAUSENS Project (P131945) will close in December 2016, the subcomponent will include financing for key nutrition interventions. The scale up also includes US$1.3 million for activities already included in this subcomponent of the original SSN project to rebuild basic infrastructure damaged by disasters (in a cyclone-proof way) as well as cash-for-work activities in the affected communities which provide temporary income support;

• Component 2: Strengthening Safety Net Administration, Monitoring, and Social Accountability (original US$6.5 million, additional US$3.1 million). Scale-up funding for project management

• Component 3: Building the Institutional Capacity for Coordination, Monitoring, and Evaluation of the Social Protection System (original US$3.3 million, additional US$1 million). Scale-up to strengthen the Government’s presence in the South of Madagascar

24. Subcomponent 1.3 (Early Recovery Response to Natural Disasters) will scale up the following activities:

(a) The payment of cash transfers to about 65,000 drought-affected households in the selected five districts. Based on updated beneficiary lists of the nutrition program, the FID as the implementing agency for all cash transfers and livelihood grants, will enroll eligible households for the cash transfer program in October 2016 so that the first cash payments can start in December 2016 to about 45,000 pre-identified households with children ages 0 to 5 years. Monthly payments will be MGA 30,000 (about US$10) per household. An additional 20,000 households with children ages 6 to 12 years will be enrolled in 2017 to receive payments as soon as possible. All cash payments will be through third-party entities such as microfinance institutions, mobile banking agents, or NGOs (in less accessible and insecure areas). In the second year, the cash transfers will evolve into CCTs linked to health and/or education services where available (for more detail, see annex II) and will continue over a two-year period. To harmonize the emergency cash program with the regular Human Development Cash Transfer (HDCT) Program implemented elsewhere in Madagascar (see annex V), the cash transfer will be reduced to the level provided under the ongoing HDCT Program (up to MGA 20,000 per month) with bimonthly payments to the mother of the household.

(b) Beneficiary households will also receive livelihood recovery grants to allow for the purchase of inputs/assets for livelihood recovery. The first livelihood grants will

9 See annex IV for project costs.

13

be made available during the first three months of the emergency response (about US$30 during December 2016 to February 2017) because of the fact that it coincides with the planting season in the South, and that the precipitation forecast for the coming season is favorable. The FAO, USAID, and others are working to ensure the provision of seeds. In 2017, a second and larger livelihood recovery grant will be provided to beneficiaries, based on some orientation in business development and as far as possible in collaboration with microfinance organizations. Beneficiaries will be assisted, over several months, by local organizations to follow through on the livelihood activities.

(c) Funding for emergency response in light of an expected strong cyclone season due to La Niña in 2017. The AF will also include up to US$1.3 million to prepare for emergency response to the expected effects (floods, cyclones, and so on) of La Niña. Activities include, as previously mentioned under Subcomponent 1.3, cash-for-work activities as well as the rehabilitation of damaged small-scale social infrastructure. FID’s existing Emergency Response Manual will be applied to these activities.

(d) Support for community-based nutrition services for treatment and prevention to approximately 60,900 children under five years of age and expanding access to 19,000 pregnant/lactating women. During the first year of the project, nutrition services will focus primarily on providing a treatment and management package for MAM. This includes screening of children, messaging on MAM to primary caregivers, four two-month treatment cycles for children under five years of age who have MAM which includes lipid-based food supplementation (Plumpy’Sup) for the child, a protective emergency ration for the family to ensure compliance with the MAM treatment,10 and household-level follow-up of families with cases of MAM. In addition, there will be referrals from the community nutrition sites to health facilities for treatment of SAM cases. In the second and third years, treatment of MAM cases will focus on a screening of children, messaging on MAM to primary caregivers, lipid-based food supplementation for the child, and household-level follow-up of families with cases of MAM. A more comprehensive nutrition package will gradually be scaled up to include preventative services. This package will complement with the needs in other parts of the country and is currently being refined as part of the preparation of the new health, nutrition, and population IDA operation (for more detail, see annex II).

10 The current national protocol for the treatment and management of MAM in an emergency context in Madagascar consists of daily intake of 92 g of Plumpy’Sup per child for two continuous months through four distributions, accompanied by a protective emergency family food ration of cereals, dried pulses, and oil (when possible). The established multisectoral nutrition cluster in Madagascar (ONN, MoH, MPSPPW, UNICEF, FAO, WHO, World Bank, WFP, and NGOs) designed this protocol because Plumpy’Sup was being shared across the family and there was no compliance with nor a significant effect of the treatment. Data from the South (2014–2016) shows that this protocol has an impact in the region. This is a compliance measure to the treatment of MAM given at the same time as Plumpy’Sup.

14

25. The World Bank team has been supporting the FID, ONN, and the MPSPPW in developing an integrated emergency response with important synergies at the household level as well as for project implementation:

• Geographic targeting. A working group composed of representatives from the MPSPPW, ONN/PNNC, FID, INSTAT, UNICEF, and the World Bank decided on geographic targeting using the latest data on malnutrition rates and food insecurity from the region to identify objectively the most drought-affected areas based on food security and malnutrition assessments by WFP and UNICEF. Additional criteria included the existence of other projects/programs, accessibility, existence of nutrition services, and the security situation. Based on this analysis, five (Tsihombe, Beloha, Ambovombe, Amboasary, and Atsimo Bekily) of the affected eight districts that belong to the two regions of Anosy and Androy were selected. Within these districts, the AF will intervene in those villages covered by the ONN’s PNNC.

• Household targeting. All households that were registered by the ONN in June 2016 as participants in the nutrition sites of the selected geographic areas (approximately 46,20011 households) are eligible for cash transfers. Of these, about 45,000 households are expected to be enrolled in the first year of project implementation. In addition, the program will enroll another 20,000 households with children ages 6 to 12 years at the end of 2017, based on a community selection process.

• Cash payments will be made through third-party institutions (mobile banking, microfinance, and/or NGOs, in insecure areas) at the community nutrition sites, as feasible, and through

11 Nearly 46,200 households are beneficiaries of the community nutrition site in the selected geographic areas and thus they are pre-identified as potential beneficiaries. Based on this list of beneficiaries, the FID will enroll eligible beneficiaries for the cash transfer program. However, not all households might show up for the enrollment process, as there may also be some double counting in the pre-identified lists, and so on, so the final number of registered beneficiaries to receive cash transfers is estimated to be only 45,000 in the first year.

Box 1: Madagascar’s National Community Nutrition Program The Government’s flagship PNNC started as a small effort in the 1990s to protect the population from the negative impacts of periods of food insecurity. Today, through a network of about 7,000 rural sites, the PNNC is now reaching 2.4 million poor mothers and children under five years of age. The program is an operational unit of the country’s ONN, charged with the overall mandate to coordinate the multi-sectoral nutrition agenda. The results from the multi-round impact evaluation (1999, 2003, 2007, and 2011) showed that the PNNC had a significant impact when it comes to improving key outcomes related to acute malnutrition, for the target population. Since 2012, recognizing the high levels of stunting in the country, the PNNC has expanded its scope to start delivering interventions to reduce chronic malnutrition. The sites, run by nutrition agents, chosen by their communities, also serve as critical links/delivery platforms for a number of different programs at the village level including for health, agriculture, education, social protection, and water and sanitation. In addition to regional-level supervision, locally contracted NGOs support the community nutrition agents (ACNs) to ensure quality service delivery. This model of service delivery has shown to be resilient against exogenous shocks and, among other things, has been useful in scaling emergency response in various parts of the country. In the South, the PNNC sites have quickly been mobilized to identify beneficiaries, screen and treat children for malnutrition, and provide an existing, pro-poor structure from which other programs can build. (For more information on the PNNC, see annex VI).

15

existing mobile banking service providers and/or local NGOs based on already established procedures and contracts.

• Complementary nutrition treatment and prevention services will be provided concurrently to the same households receiving the cash transfers through scaling-up an existing network of community nutrition sites that have proven to be effective in addressing acute malnutrition. Supervision and monitoring of nutrition activities will be provided by already established regional nutrition officers and local NGOs contracted by the U-PNNC.

• A joint communication strategy is being developed by the implementing agencies (FID, ONN/U-PNNC, and MPSPPW) to foster transparency and encourage families to focus on their children’s development as well as their productive recovery

• A coordinated complaint mechanism will be introduced for beneficiaries and other citizens to alert the project implementing agencies of any bottlenecks and wrongdoings. This complaint mechanism is based on the procedures that FID is already implementing under the other safety net programs and will be gradually expanded to cover complaints pertaining to the nutrition services. Details are described in the respective FID Emergency Cash Transfer Operational Manual as well in the respective PNNC Manual.

• Coordination at national and local levels. The MPSPPW in coordination with the BNGRC is responsible for the overall coordination of the country’s response to the drought in the South and has already hosted a number of coordination workshops and high-level meetings to enhance a coordinated approach. At the local level, the humanitarian agencies under the leadership of BNGRC’s local offices, are already meeting regularly at the district and regional levels. This local coordination mechanism already includes both the FID and the ONN. In addition, the MPSPPW will strengthen its presence in the South, with support from the AF, to foster the coordination of relevant government agencies with the humanitarian and development partners in this region.

26. Components 2 and 3 of the SSNP (Strengthening Safety Net Administration, Monitoring, and Social Accountability Building; and the Institutional Capacity for Coordination, Monitoring, and Evaluation of the Social Protection System) will also be scaled up under the AF. US$3.1 million will be added to Component 2 of the SSNP to ensure proper management of the safety net activities under the AF. This will include the opening of an additional FID office in the South (Fort Dauphin)12 and a satellite office in Ambovombe, as well as funding for logistics, training of trainers, communication, and so on. Furthermore, Component 3 will get an additional US$1 million to enable the MPSPPW to carry out its coordination, monitoring, and evaluation functions in the South. This includes the rehabilitation of the two local offices, provision of transport, operating costs, and technical assistance to enhance the Ministry’s presence in the region and facilitate coordination, particularly between the humanitarian and the development partners. It also includes funding for program monitoring and evaluation. Both components include funding for small works for the upkeep and maintenance of the project implementation agencies’ premises; which is covered under the Project’s Environmental and

12 Fort Dauphin was selected due to its connectivity, available office space, and coordination platform with multiple partners.

16

Social Safeguards Management Framework (ESMF). The ESMF of the original financing has been revised and updated to cover this AF. See annex IV for project costs.

27. Implementation arrangements. Subcomponent 1.3 and Component 2 are implemented by the Government of Madagascar’s FID. The nutrition activities that will be included under Subcomponent 1.3 as part of the AF (for a total amount of US$8.5 million) will be implemented by the U-PNNC under the leadership of the ONN. The same U-PNNC manages the implementation of the nutrition component of the PAUSENS Project. Component 3 continues to be implemented by the MPSPPW, as under the SSNP. As such, all three institutions will carry the fiduciary responsibilities for their respective activities. A Memorandum of Understanding between the FID, ONN, and the MPSPWP will outline the roles and responsibilities of the involved institutions, as well as the mechanisms for coordination, communication, and monitoring. In addition, all fiduciary aspects of the program will adjusted to account for the new implementation agency.

28. Alternatives considered. During preparation of the AF for the SSNP, the Government and the World Bank team discussed several options for project implementation including a wider geographical coverage (that is, covering all eight affected districts) while shortening the time of the intervention (to about a year). This option was discarded because not all districts are equally affected, other humanitarian and development partners also provide support, and a shorter duration of the support might jeopardize the recovery and resilience objective of the AF. Another alternative considered consisted of the ‘deeper’ coverage of the population in the most affected five districts by covering more villages than those that have an existing nutrition site in these districts, and also over a shorter period of time. This option was not chosen due to the time it would have taken to select and register the population as they are not yet registered with the ONN/PNNC. The proposed approach seeks to balance a rapid response (based on the ONN/PNNC’s existing beneficiary lists), the coverage of the districts with the highest malnutrition rates (five instead of eight districts), and the population’s need for short-term assistance well as medium-term livelihood recovery support.

III. Proposed Changes

Summary of Proposed Changes The proposed AF will provide short- to medium-term assistance to about an additional 65,000 households in five districts of Madagascar’s Southern region (see map in annex III) through (a) the payment of 36 months of cash transfers to households with children ages 0 to 5 years (who have been registered with the PNNC) and additional households with children ages 6 to 12 years in the second year; (b) livelihood recovery grants for households whose use will be prepared through community-based training; and (c) nutrition treatment and prevention services throughout the 36 months.

The nutrition services are new to the project as part of the AF and the U-PNNC will be added as a new implementing agency to manage these activities.

The PDO and the components will remain the same. The results framework has been updated to reflect the additional beneficiaries and new indicators have been added at PDO and intermediate levels to reflect the objectives of the AF. Finally, project costs, disbursement estimates, and FM arrangements have been adjusted accordingly.

Change in Implementing Agency Yes [ X ] No [ ]

Change in Project's Development Objectives Yes [ ] No [ X ]

17

Change in Results Framework Yes [ X ] No [ ]

Change in Safeguard Policies Triggered Yes [ ] No [ X ]

Change of EA category Yes [ ] No [ X ]

Other Changes to Safeguards Yes [ ] No [ X ]

Change in Legal Covenants Yes [ ] No [ X ]

Change in Loan Closing Date(s) Yes [ ] No [ X ]

Cancellations Proposed Yes [ ] No [ X ]

Change in Disbursement Arrangements Yes [ X ] No [ ]

Reallocation between Disbursement Categories Yes [ ] No [ X ]

Change in Disbursement Estimates Yes [ X ] No [ ]

Change to Components and Cost Yes [ X ] No [ ]

Change in Institutional Arrangements Yes [ X ] No [ ]

Change in Financial Management Yes [ X ] No [ ]

Change in Procurement Yes [ X ] No [ ]

Change in Implementation Schedule Yes [ ] No [ X ]

Other Change(s) Yes [ ] No [ X ]

Development Objective/Results PHHHDO Project’s Development Objectives Original PDO The PDO is to support the Government in increasing the access of extremely poor households to safety net services and in laying the foundations for a social protection system.

Change in Results Framework PHHCRF Explanation: The additional beneficiaries and new activities supported by the AF are reflected in the results framework. A set of PDO-level and intermediate indicators have been added to monitor the progress of the three main activities proposed under the AF. See annex I for details.

Compliance

Covenants - Additional Financing ( AF Social Safety Net - P160554 ) Source of Funds

Finance Agreement Reference

Description of Covenants Date Due Recurrent Frequency Action

IDAT Schedule 2, section 1, B 1 (b)

The Recipient and the Project Implementing Agency [FID]

New

18

shall, not later than 30 days after the Effective Date, amend the Subsidiary Agreement to include the amount of the Additional Financing as well as the activities, which have been added to the Original Project.

Conditions PHCondTbl

Source Of Fund Name Type IDAT Signed U-PNNC Subsidiary

Agreement Effectiveness

Description of Condition The UPNNC Subsidiary Agreement has been executed on behalf of the Recipient and UPNNC.

Risk PHHHRISKS Risk Category Rating (H, S, M, L) 1. Political and Governance High 2. Macroeconomic Moderate 3. Sector Strategies and Policies Substantial 4. Technical Design of Project or Program Substantial 5. Institutional Capacity for Implementation and Sustainability Substantial 6. Fiduciary Substantial 7. Environment and Social Moderate 8. Stakeholders Moderate 9. Other OVERALL Substantial

Finance Loan Closing Date - AF Social Safety Net (P160554 )

Source of Funds Proposed Additional Financing Loan Closing Date IDA Grant 30-Sep-2020

Change in Disbursement Arrangements PHHCDA

19

Explanation: The SSNP uses Interim Financial Reports for disbursements. So far, the U-PNNC has used Statement of Expenditure-based disbursement procedures only. Therefore, the project will partially disburse based on Interim Financial Reports (FID, safety net activities) and partially based on Statement of Expenditures (U-PNNC, nutrition service activities). In addition to the two existing designated accounts of the SSNP, the U-PNNC will open a designated account to manage the nutrition activities.

Change in Disbursement estimates (including all sources of Financing)PHHCDE Explanation: The US$35 million AF will be disbursed over a period of three years and as reflected in the following table.

Expected Disbursements (in US$, millions)(including all Sources of Financing) Fiscal Year 2016 2017 2018 2019 2020 2021 2022 2023 2024 2025 Annual 1.54 23.18 22.78 17.92 7.78 1.80 0.00 0.00 0.00 0.00 Cumulative 1.54 24.72 47.50 65.42 73.20 75.00 0.00 0.00 0.00 0.00

Allocations - AF Social Safety Net (P160554)

Source of Fund Currency Category of

Expenditure Allocation Disbursement %(Type

Total)

Proposed Proposed

IDA Grant US$ 15,000,000.00 100.00

Total: 15,000,000.00

IDA Grant from CRW US$ 20,000,000.00 100.00

Total: 20,000,000.00

Components Change to Components and Cost PHHCCC Explanation: The AF will increase the total costs of Components 1 (additional US$30.9 million), 2 (additional US$3.1 million), and 3 (additional US$1.0 million). The changes are summarized in the following table.

Current Component Name

Proposed Component Name

Current Cost (US$,

millions)

Proposed Cost (US$, millions)

Action

Component 1: Building a Safety Net for the Poor in Selected Rural Areas

Component 1: Building a Safety Net for the Poor in Selected Rural Areas

30.20 61.10 Revised

Component 2: Strengthening Safety Net Administration, Monitoring, and Social Accountability

Component 2: Strengthening Safety Net Administration, Monitoring, and Social Accountability

6.50 9.60 Revised

20

Component 3: Building the Institutional Capacity for Coordination, Monitoring and Evaluation of the Social Protection System

Component 3: Building the Institutional Capacity for Coordination, Monitoring and Evaluation of the Social Protection System

3.30 4.30 Revised

Total: 40.00 75.00

Other Change(s) Change in Implementing Agency PHH Explanation: The U-PNNC, currently one of the implementing agencies of the PAUSENS Project (P131945), has been added under the AF as the implementing agency for the nutrition-related activities. The ONN/U-PNNC is a long-term implementing partner for the World Bank and has a strong track record of satisfactory implementation. Additionally, the ONN/U-PNNC has been implementing emergency nutrition interventions in the South over the past year with financing from the PAUSENS Project.

Implementing Agency Name Type Action FID Implementing Agency No Change U-PNNC Implementing Agency New MPSPPW Implementing Agency No Change

Change in Institutional Arrangements Explanation: The U-PNNC will be responsible for the implementation of the complementary nutrition services under the AF. The implementing agencies of the SSNP (FID and the MPSPPW) have signed a Memorandum of Understanding with the ONN/U-PNNC to outline the respective roles and responsibilities as well as their coordination arrangements, satisfactory to the World Bank and before negotiations.

Change in Financial Management PHHCFM Explanation: Besides the existing Interim Financial Reports, the SSNP will now also have arrangements for including Statement of Expenditure-based disbursements for the U-PNNC. See FM assessment in annex II.

Change in Procurement Explanation: A procurement capacity assessment of the U-PNNC was carried out because it is a new implementing agency under this project. Because the initial credit of the PAUSENS will close in December 2016, only a few procurement activities remain to be followed up by the U-PNNC. The U-PNNC currently has a procurement officer and two procurement assistants, which is sufficient capacity. As such, no additional staff is required to carry out procurement activities under the proposed AF. The central office will initiate all procurement activities as well as update and refine the approved procurement plan by project effectiveness. See procurement assessment in annex II.

21

IV. Appraisal Summary

Economic and Financial Analysis PHHASEFA

Explanation: The economic and financial analysis of the SSNP and the PAUSENS Projects remain relevant. The AF will further contribute toward the well-being of households and communities providing predictable safety net transfers, nutrition services, and a livelihood recovery grant to cope sustainably with the devastating effects of the drought. The World Bank’s added value through the AF consists of providing global knowledge and experience in adaptive safety nets, community based nutrition services and support for productive recovery in the aftermath of disasters including droughts, floods and man-made shocks. It also fosters donor coordination and harmonization across programs and interventions, and promotes a strategic approach as well as a pragmatic response to climate shocks. The following analysis provides a discussion of different options and experiences building on global knowledge.

Economic analysis for the cash transfers under the AF

This economic analysis provides an ex ante estimate of the AF’s potential impact on poverty and consumption for different benefit scenarios. The main source of data for this analysis is Enquête national sur le suivi des objectifs du millénaire pour le développement à Madagascar carried out by INSTAT in 2014. As the data were collected before the current drought, it is likely that the country has experienced negative changes in consumption at the household level, thus this analysis likely underestimates the impact that this program would have. The analysis is limited to the two regions where the program will be taking place: Androy and Anosy. These regions have the five districts with the highest degree of malnutrition. For estimates of the costs in U.S. dollars, an exchange rate of MGA 3050 per U.S. dollar is used.

The benchmark benefit scenario considered in this analysis is an unconditional cash transfer of MGA 30,000 per month, per household for a period of 12 months for 45,000 households, followed by a CCT of MGA 20,000 per month for 24 months (bringing an additional 20,000 households into the program) along the lines of the HDCT Program. The original 45,000 beneficiaries will also receive a livelihood recovery grant of MGA 180,000 in total during the first 12 months, to rebuild their income generating activities.

Table 4.1 (see annex IV) summarizes the estimated impact on poverty and consumption of the benefits provided through the cash transfer program. It explores variations of benefit packages, beneficiaries, and program length while remaining approximately budget neutral, to highlight the trade-offs between the degree of coverage that the program can afford and the impact on poverty and consumption.

Column 1 shows that the benchmark scenario of MGA 30,000 of an unconditional cash transfer for 12 months, followed by 24 months of a CCT will increase consumption of the average household by 67 percent, 13 and cover 32 percent of poor households in the districts covered. The program will cover 16 percent of the food poverty gap in this area.

Columns 2–4 vary the size of the monthly transfer amount, the number of beneficiary households, and the months of coverage.

o In column 2, increasing the length of the cash transfer program by six months and decreasing the CCT by six months, reduces the number of beneficiaries who are added into the program in the second year from 20,000 to 19,200 beneficiaries, but increases annual food consumption for the average household beneficiary from 67 percent to 72 percent.

o In column 3, leaving the benefit size unchanged at MGA 30,000 for the entire length of the 13 The impact on consumption is rather high at 67 percent because the food poverty gap index (the extent to which individuals fall below the food poverty line as a proportion of the food poverty line) is itself high. Thus, this project has a substantial potential impact on increasing the consumption of poor households.

22

program, reduces the number of beneficiaries possible in the expansion in the second year from the benchmark of 20,000 to 16,200 and the program length by seven months. However, it increases annual food consumption for the average household beneficiary from 67 percent to 86 percent.

o Column 4 shows that decreasing the benefit size to MGA 20,000 for the length of the program increases the number of beneficiaries possible in the expansion from 20,000 to 31,250, thus increasing the coverage of food-poor households from 32 percent to 37 percent. However, the program only contributes to an increase in the annual food consumption for the average household beneficiary of 57 percent.

In summary, given the objectives of the program to deliver benefits to the poorest households, the economic analysis of the cash transfers shows that the proposed project should implement the benchmark scenario: MGA 30,000 per month, per household for a period of 12 months, followed by a CCT for MGA 20,000 per month for 24 months along the lines of the HDCT Program. The other scenarios that were explored sacrifice additional coverage for either increased benefit sizes and/or extended program length. The benchmark scenario presents an ideal way to reach a large number of beneficiaries, significantly increasing their consumption in the first year, and allows the program to expand in the second and third years to additional beneficiaries, maximizing the impact of the program.

Economic analysis for the AF nutrition activities

Given that increased coverage for a set of direct nutrition interventions will be achieved through the proposed AF, the overall economic impact is expected to be as great as or greater than the original project. According to the latest evidence, nutrition interventions are more effective when they target the poorest segments of the population. The proposed AF focuses on those geographic regions of Madagascar that have the highest incidence of food poverty and malnutrition. Its population is often unable to articulate demand for services, as well as bear the economic costs associated with using those services (transport, fees for services, and so on). Therefore, there is a strong justification for public sector involvement through a combination of demand- and supply-side interventions such as cash transfers (to enable demand) and nutrition services (to strengthen supply).

In 2008, the Copenhagen Consensus ranked nutrition interventions as providing some of the most effective returns of all development interventions. Moreover, The Lancet, one of the world's most respected medical journals, examined evidence from hundreds of studies in a variety of country settings and identified 13 highly cost-effective direct interventions such as vitamin A and zinc supplements, iodized salt, and the promotion of healthy behaviors including handwashing, exclusive breastfeeding, and complementary and therapeutic feeding practices. In line with these findings, the AF is financing a subset of nutrition interventions (phased in by starting with treatment and management of acute malnutrition) for which there are existing delivery mechanisms as advised in the publication: Scaling-Up Nutrition, What Will it Cost? The nutritional interventions valued at US$8.5 million for the duration of the project, will support interventions to increase access to direct nutrition interventions for acute and chronic malnutrition to approximately 61,000 children under five years benefitting from nutrition services, including 38,000 children treated for MAM. In addition, approximately 19,000 pregnant/lactating women will have access to enhanced nutrition services.

Nutrition services have been shown to have a very favorable cost-benefit ratio with regard to short- and long-term health benefits as well as on education achievements and, last but not least, on productivity outcomes. Investing in good infant and child nutrition leads to an estimated 2 percent to 3 percent growth in the economic wealth of developing countries. Further, costs of recommended nutrition-specific interventions have been compared with the economic value of their benefits (with regard to reduced mortality and morbidity) and the basic benefit-to-cost ratio is a very high 15.8 to 1. Even the low case, with extremely conservative estimates, yields a benefit-to-cost ratio of 4.8 to 1, and the high case value is

23

40 to 1, in line with the Copenhagen Consensus. Research has also indicated that height-for-age at two years was the best predictor of future productivity. In the case of the South, where the population suffers from the dual burden of acute and chronic malnutrition impacts, the benefits of investing in this area will likely be exponential. Additionally, providing nutrition services is an important preventive safety net for the population in this region.

Technical Analysis PHHASTA

Explanation: For the existing social safety net activities under the original SSNP, the AF does not alter the Technical Analysis of the appraisal summary. The following technical aspects are noteworthy related to the proposed AF:

Geographical targeting. A working group composed of representatives from the MPSPPW, ONN/PNNC, FID, INSTAT, UNICEF, and the World Bank, decided the geographical targeting, using the latest data on malnutrition rates from the region to objectively identify the most affected areas. Other criteria included the existence of other projects/programs, the accessibility, existence of nutrition services, and the security situation. Based on this analysis, the Government selected five districts (Tsihombe, Beloha, Ambovombe, Amboasary, and Atsimo Bekily) in two regions of the Deep South (Anosy, Androy).

Household targeting. The selection of safety net beneficiary households will be done based on the existing lists of participants in the nutrition service. The ONN sites have detailed lists of mothers with children under six years that they keep updated. These lists are presently updated and digitalized so that they can be used immediately for safety net registration.

Beneficiary registration for the safety net will be carried out by the FID at the nutrition sites. The FID will contract experienced agencies that can quickly register the listed beneficiaries, provide them with information on the program, and assist the beneficiaries throughout the implementation of the program. They will be present on payment day, help with complaints and updates (for example, if a beneficiary leaves the area, she can be replaced by another mother with a newborn child), and prepare the beneficiaries for the receipt of the livelihood grant.

Cash payments will be made monthly to the mothers, at the ONN nutrition site as far as possible. The payment will be provided through service providers and priority will be given to established payment services that already have experience in making cash payments. For areas in which no such agency can be found, the FID would contract NGOs to provide the monthly payments.

Safety net complementary measures. International experience shows that providing extremely poor families with cash is an important instrument to reach immediate effects such as increased consumption. However, to promote the early recovery of livelihoods, it will be necessary to accompany the cash transfers with guidance for beneficiaries on, for example, financial planning, options for productive activities, group organization, citizen engagement, self-affirmation, and other subjects related to the well-being and prospective development of the households. These activities will be developed during the first months of project implementation to kick off during the second year of the program.

The nutrition interventions are building on extensive past and current experiences in Madagascar on nutrition across the country and in the South. As such, the interventions envisaged are based on well-known and tested approaches that will not add to the overall complexity of the project. The focus will be on gradually phasing in a scale-up of proven cost-effective interventions starting with a focus on increasing coverage of MAM treatment and management and then moving to provide access to a more comprehensive nutrition package which includes preventative nutrition services and key complementary activities to promote better nutrition to the target population. The interventions will largely be delivered through community nutrition sites under the PNNC. In Madagascar, this community-based nutrition approach has proven to be very effective in reducing rates of acute malnutrition. Results from the multi-round

24

anthropometric studies conducted on the program in 1998, 2004, 2007, and 2011 show that acute malnutrition progressively declined by more than half in the under-three age group (from 36.1 percent to 17.8 percent) among children who were enrolled in the PNNC as compared to insignificant changes among communities where children were not enrolled in the program.

Contracting of service providers. The capacity of service providers in Madagascar, including in the South, is generally considered weak. Given the necessity to work through local and national service providers in the implementation of the AF (for example, for managing the enrollment of beneficiaries, for payments, for training and supervision of nutrition services, and for any complementary measures provided under the AF), the AF promotes a two-pronged approach. As far as possible, experienced service providers (for example, in microfinance and/or mobile banking institutions) will also be contracted, through a competitive process, to provide services in the South. Furthermore, contracts will be crafted in a results-based fashion. Finally, close monitoring of the quality of service provision will be coordinated between the three implementing agencies.

Communication will be important, not only to inform the general public and the beneficiaries of the program and its social accountability, but also to transmit important messages regarding the parents’ co-responsibility for their children’s development (nutrition, school attendance, parenting, health/vaccination, and so on), financial planning, and productive investments. These will be intensified over time to ensure that the program produces a value added beyond the smoothing of consumption (through regular cash transfers) and the availability of supply-side nutrition services. The implementing agencies have developed a joint communications strategy for the proposed AF, which includes innovative elements such as interactive theatre, radio shows, and other cultural activities that would motivate and empower the beneficiaries to become active agents of their well-being.

Controls and accountability. Similar mechanisms to those developed for the other safety net programs under the SSNP will be implemented. Given the remoteness of the southern regions, their difficult security situation, and the lack of experience with cash transfers in these areas, the FID and ONN/PNNC will need to monitor the program implementation closely and coordinate their supervision activities. As is usual, complaint and grievance redress mechanisms will be set up to ensure that the beneficiaries have channels for eventual grievances. In addition, the FID will ensure intensive communication and information about the program implementation features so that beneficiaries and other stakeholders know the rights and responsibilities of participating in the program.

Social Analysis PHHASSA

Explanation: The AF does not change this part of the appraisal summary.

A recent gender analysis carried out by a European Union-financed project, with support from German Agency for International Cooperation (Deutsche Gesellschaft fur Internationale Zusammenarbeit), finds that 78.35 percent of women in the southern region are in a precarious situation. Only 21.65 percent of rural women are legally married, 48 percent live in a traditional marriage, 13.5 percent are separated, 8 percent are widows, 5 percent live in concubinage, 3.8 percent are single, and others did not report. The study also shows that ultimately, women’s access to and control of the region’s main household resources —land and zebu—are entirely dependent on the position of men, of which some are fiercely opposed to women becoming owners, and some are more flexible and give the women a small share of the property. Due to this unequal ownership of assets, access to microfinance is also extremely restricted for women. In addition, given the extremely high illiteracy rate, especially among women (for example, in Anosy, the male illiteracy rate is 54 percent and the female illiteracy rate is 71 percent), it comes as no surprise that women are extremely limited in income generation and that their productive activities depend heavily on the men who may or may not be present given the region’s partial polygamy culture. Interestingly, the study asked women what they needed most for their productive activities; they said they primarily needed

25

money for their agricultural work, for example small tools as necessary to water the vegetable gardens, a plow, and even small livestock. The vast majority of women interviewed had no clear idea for a personal or group ‘project’ they wished to carry out to improve their livelihoods and increase their incomes. This is an important aspect that the proposed AF will need to address through training sessions and behavioral interventions.

Environmental Analysis

Explanation: The SSN project has been categorized as “B” in relation to social and environmental safeguards and has triggered three safeguard categories: Environmental Assessment (OP/BP 4.01), Physical Cultural Resources (OP/BP 4.11) and the Involuntary Resettlement Policy (OP/BP 4.12). Considering the expected positive impacts of the project on the environment and gender, as well as its potentially minor negative impacts through, for example, small scale infrastructure rehabilitation in the aftermath of disasters, the risk rating for safeguards is rated as moderate. The Additional Financing does not change this part of the Appraisal Summary. The proposed AF is modest in scope and the safeguards category “B” rating will remain unchanged. No new environmental risks are expected to arise and the AF is not foreseen to trigger any new safeguard policies. For this AF, only the activity 2 under the nutrition services on the distribution of lipid-based food supplementation (Plumpy’Sup) to children could present a potential risk of inconsumable products after the products’ expiration date, or damaged products. The revised ESMF includes appropriate elimination procedures of expired or damaged products. The project’s ESMF and Involuntary Resettlement Framework were revised and updated to cover this AF. Prior to negotiations, these documents were approved by the Bank as well as disclosed in the country and through the Infoshop on September 30, 2016.

Risk

Explanation: The overall risk remains Substantial. The Political and Governance risk remains high as the country still emerges from the recent political and economic crisis that has put to the test many of its institutions for good governance and accountability. The risks for Sector Strategies and Policies, Technical Design of Project, and the Fiduciary risk remain substantial and mitigation measures from the original Project continue to be implemented through support to policy development, technical assistance and fiduciary controls. In addition, the risk for Institutional Capacity for Implementation has increased to substantial given the inclusion of an additional implementation agency in the project. This risk is being mitigated through a tripartite Memorandum of Understanding, which outlines the roles and responsibilities of each of the three implementing agencies under the project, as well as the joint development of the operational manuals for the AF as well as a joint evaluation agenda.

V. World Bank Grievance Redress

29. Communities and individuals who believe that they are adversely affected by a World Bank (WB) supported project may submit complaints to existing project-level grievance redress mechanisms or the WB’s Grievance Redress Service (GRS). The GRS ensures that complaints received are promptly reviewed in order to address project-related concerns. Project affected communities and individuals may submit their complaint to the WB’s independent Inspection Panel which determines whether harm occurred, or could occur, as a result of WB non-compliance with its policies and procedures. Complaints may be submitted at any time after concerns have been brought directly to the World Bank's attention, and Bank Management has been given an

26

opportunity to respond. For information on how to submit complaints to the World Bank’s corporate Grievance Redress Service (GRS), please visit http://www.worldbank.org/GRS. For information on how to submit complaints to the World Bank Inspection Panel, please visit www.inspectionpanel.org.

27

Annex I: Revised Results Framework and Monitoring

Project Development Objectives Original Project Development Objective - Parent: The project development objective (PDO) is to support the government in increasing the access of extremely poor households to safety net services and in laying the foundations for a social protection system.

Results Core sector indicators are considered: Yes Results reporting level: Project Level .

Project Development Objective Indicators Status Indicator Name Core Unit of Measure Baseline Actual(Current) End Target Revised Direct project beneficiaries Number Value 0.00 23691.00 863500.00

Date 01-Jul-2015 30-Sep-2016 30-Sep-2020

Comment This includes: - 547,500 original, - 300,000 safety net, and - 16,000 additional nutrition

No Change Female beneficiaries Percentage Value 0.00 60.00 65.00

Sub Type Supplemental

No Change Percentage of beneficiaries of the safety net programs (HDCT, PSN) belonging to the 30% poorest population.

Percentage Value 0.00 100.00 70.00

Date 01-Jul-2015 30-Sep-2016 30-Sep-2020

Comment

28

No Change Percentage of primary school aged children of HDCT beneficiary families who attend school at least 80% of the time

Percentage Value 0.00 90.00 80.00

Date 01-Jul-2015 30-Sep-2016 30-Sep-2020

Comment No Change Integrated beneficiary registry

established at MPSPWP Yes/No Value No No Yes

Date 01-Jul-2015 30-Sep-2016 30-Sep-2020

Comment New Pregnant/lactating women,

adolescent girls and/or children under age five-reached by basic nutrition services (number)

Number Value 0.00 0.00 60900.00

Date 30-Sep-2015 30-Sep-2016 30-Sep-2020

Comment New Children under age five treated

for moderate or severe acute malnutrition (number)

Number Value 0.00 0.00 38400.00

Sub Type Date 30-Sep-2016 30-Sep-2016 30-Sep-2020

Breakdown Comment New Children under the age of 24

months benefiting from improved infant and young child feeding (IYCF) practices

Number Value 0.00 22500.00

Sub Type Date 30-Sep-2016 30-Sep-2020

Breakdown Comment Intermediate Results Indicators Status Indicator Name Core Unit of Measure Baseline Actual(Current) End Target New Percentage of safety net

beneficiaries in the South using the livelihood grant for productive activities

Percentage Value 0.00 0.00 80.00

Date 30-Sep-2016 30-Sep-2016 30-Sep-2020

Comment New Children 0 to 59 months

monitored at the community nutrition sites

Number Value 0.00 0.00 50700.00

Date 30-Sep-2016 30-Sep-2016 30-Sep-2020

Comment

29

Revised Beneficiaries of Safety Nets programs (number) Number Value 5000.00 6300.00 172500.00

Date 01-Jul-2015 30-Sep-2016 30-Sep-2020

Comment Added an additional 63,000 beneficiaries of additional financing

No Change Beneficiaries of Safety Nets programs - Conditional cash transfers (number)

Number Value 5000.00 6300.00 39000.00

Sub Type Date 01-Jul-2015 30-Sep-2016 30-Sep-2020

Breakdown Comment Number of Households. Multiply by 5 to get number of people

Revised Beneficiaries of Safety Nets programs - Female (number) Number Value 4000.00 5300.00 133000.00

Sub Type Date 01-Jul-2015 30-Sep-2016 30-Sep-2020

Breakdown Comment CCT: 38000 women PSNP: 16,000 women AF CT: 60,000 Disaster: 19000

No Change Beneficiaries of Safety Nets programs - Cash-for-work, food-for-work and public works (number)

Number Value 0.00 3900.00 70500.00

Sub Type Date 01-Jul-2015 30-Sep-2016 30-Sep-2020

Breakdown Comment No Change Number of landscape

development plans developed in a participatory way

Number Value 0.00 18.00 60.00

Date 01-Jul-2015 30-Sep-2016 30-Sep-2020

Comment

30

No Change Area treated with SWC measures Hectare(Ha) Value 0.00 0.00 900.00

Date 01-Jul-2015 30-Sep-2016 30-Sep-2020

Comment No Change Area re/afforested through

productive safety net activities Hectare(Ha) Value 0.00 0.00 540.00

Date 01-Jul-2015 30-Sep-2016 30-Sep-2020

Comment No Change Number of employment days

created under the PSN and disaster response Cash for Work activities

Number Value 0.00 78000.00 8560000.00

Date 01-Jul-2015 30-Sep-2016 30-Sep-2020

Comment No Change Percentage of registered

complaints that have been addressed by FID

Percentage Value 0.00 98.00 90.00

Date 01-Jul-2015 30-Sep-2016 30-Sep-2020

Comment No Change Percentage of community

assets maintained one year after their completion

Percentage Value 0.00 0.00 80.00

Date 01-Jul-2015 30-Sep-2016 30-Sep-2020

Comment No Change Operational costs of FID as a

percentage of program implementation expenditures

Percentage Value 20.00 20.00 16.20

Date 01-Jul-2015 30-Sep-2016 30-Sep-2020

Comment No Change Establishment of a social

protection program monitoring system at the MPSPPW

Yes/No Value No No Yes

Date 01-Jul-2015 30-Sep-2016 30-Sep-2020

Comment No Change Impact Evaluation carried out

and results available publicly Yes/No Value No No Yes

Date 01-Jul-2015 30-Sep-2016 30-Sep-2020

Comment

31

New Number of Agreements with government, non-government and development partners signed by the MPSPPW to harmonize the social sector related programs and activities in the South of Madagascar

Number Value 1.00 1.00 6.00

Date 30-Sep-2016 30-Sep-2016 30-Sep-2020

Comment

New Regional Social Protection committees established in Anosy and Androy (South) and functioning through regular meetings.

Number Value 0.00 0.00 2.00

Date 31-Jul-2016 30-Sep-2016 30-Sep-2020

Comment

.

32

Revised Results Framework and Monitoring

Revisions to the Results Framework Comments/Rationale for Change

PDO Indicators Current Proposed change*

Direct project beneficiaries End target revised Revised upward to include additional population to be targeted by the AF

New: Pregnant/lactating women, adolescent girls, and/or children under age five reached by basic nutrition services (number) Breakdown Indicators: a) Children under age five treated for

moderate or severe acute malnutrition (number)

b) Children under the age of 24 months benefiting from improved infant and young child feeding (IYCF) practices (number)

To account for new nutrition activities under the AF

Intermediate Results

Current Proposed change New: Percentage of safety net beneficiaries in

the South using the livelihood grant for productive activities.

To account for new safety nets activity in the South under the AF

New: Children ages 0 to 59 months monitored at the community nutrition sites

To account for nutrition activities under the AF

New: Number of agreements (conventions) signed with partners (government, nongovernment, and/or development partners) related to the harmonization of programs and activities in the South

To account for new coordination activities under the AF

New: Regional social protection committees in the southern regions of Androy and Anosy established and functioning.

To account for new coordination activities under the AF

Beneficiaries of safety nets programs (number) Breakdown Indicator: Beneficiaries of safety nets programs - Female (number)

End target revised Revised upward to include additional population to be targeted by the AF

Note: *Bold defines project target population.

33

Annex II: Detailed Description of Modified or New Project Activities 1. The proposed AF will support the scale-up of existing project activities, under the following components of the SSNP:

(a) Subcomponent 1.3: Early Recovery Response to Natural Disasters with cash transfers, nutrition treatment and prevention services, and livelihood recovery grants. In light of the high rate of MAM among children under five years and the fact that the initial credit of the IDA-funded PAUSENS (P131945) will close in December 2016, the subcomponent will including financing for targeted nutrition interventions.

(b) Component 2: Strengthening Safety Net Administration, Monitoring, and Social Accountability with funding for project management

(c) Component 3: Building the Institutional Capacity for Coordination, Monitoring and Evaluation of the Social Protection System with funding for strengthening the Government’s presence in the south of Madagascar.

2. The original design of Subcomponent 1.3 will continue to ‘enable the Government of Madagascar to help households to recover after a disaster by providing them with cash-for-work opportunities and by rehabilitating and reconstructing damaged infrastructure in the affected communities’. These disaster response activities have been designed to respond to rapid-onset disasters like those produced by cyclones and flooding, which usually create destruction through immediate physical impact. They strike different regions and communities in an unpredictable manner and have been more severe because of climate change in the last few years. Three new activities will be included under Subcomponent 1.3.

3. The AF will expand the scope of the subcomponent to enable the Government to respond to a different kind of disaster: a prolonged drought. In this type of slow-onset disaster, which creates longer-term social and economic crises in affected areas, it would not be sufficient to provide households with short-term cash-for-work opportunities, as envisioned under the original subcomponent in response to cyclones and flooding. The ongoing drought has deprived affected households from producing food for their daily consumption/subsistence and compromised their well-being with regard to health and nutrition outcomes. In addition, its continuation over a medium-term period has also likely led households to take children out of school (given school fees and opportunity costs) and sell assets and capital which normally they could have used to start their livelihood activities once the drought ceased. In this situation, it is essential to provide households with income and nutrition support in the short and medium term, as well as recovery support to enable them to resume their livelihood activities.

34

4. The AF provides essential support directly to households through distinct but coordinated activities:

(a) Activity 1. Cash transfers for households participating in the National Nutrition Program with children ages 0 to 5 years. All households registered into the program in the targeted areas will benefit from regular cash transfers throughout the program. However, the cash transfer program will have two phases: (i) a livelihood recovery phase during the first 10 months; and (ii) the progressive evolution of the cash transfer program into a regular CCT program implemented in other regions under the SSNP. For the first 10 months (corresponding to the early recovery phase), cash transfers of MGA 30,000 (approximately US$10) will be made every month to beneficiary households. For this first phase, the cash transfers will not be subject to compliance with any co-responsibility. The first three payments (starting in December 2016) will be double, therefore equivalent to MGA 60,000, given the households’ de-capitalization and the upcoming planting season in December 2016 to February 2017. Thereafter, during the following months, the beneficiary households will continue to receive cash transfers of MGA 30,000 per month until September 2017. As of the second year of the AF (that is, after the early recovery phase) and until the end of the third year, benefits will be aligned with those of the Human Development Cash Transfer (HDCT) program implemented elsewhere in the country (approximately MGA 20,000 per household). At that point, households with children ages between 6 and 12 years who did not participate in the first early recovery phase will be invited to join the program, and will be selected using community targeting. As of the second year, all beneficiary households will be required to comply with co-responsibilities linked to enrollment and attendance of their children in school and/or health related co-responsibilities (such as vaccination), following the same design as the HDCT—where possible. If primary education or health services are not reliably available at the village level, no co-responsibilities will be introduced, thus the cash transfers will remain unconditional.

(b) Activity 2. Nutrition services will be delivered in targeted communities through the existing PNNC’s nutrition sites (nationally known as ‘Seecaline’ sites). This is a long-standing, community-based service delivery model in the country which is currently also used in the five regions supported by the PAUSENS Project to deliver nutrition services (see annex VI for an overview of the nutrition sector in Madagascar). The nutrition sites are managed by ACNs selected by the community who are responsible for delivering a defined package of direct nutrition services. The main objective of the nutrition activities is to improve household access to basic nutrition services with prevention and treatment of malnutrition. There will be 338 sites implicated in five districts (four in the Androy Region and the Amboasary District in the Anosy Region) of which 251 sites are already operational. Given the low coverage in some areas, the ONN/U-PNNC will establish 87 new sites and the start-up costs will include selection and training of ACNs, animators, and other key local stakeholders including the community. Locally contracted NGOs and the ONN regional nutrition officers will be responsible for supervision and technical support to the community nutrition sites. During the first year of the project, nutrition services will primarily focus on providing

35

a treatment and management package for MAM.14 This includes screening of children, messaging on MAM to primary caregivers, four two-month treatment cycles for children under five years who have MAM, which includes lipid-based food supplementation for the child, a protective food ration for the family to ensure compliance of the MAM treatment for the child, and household-level follow-up of children. In addition, there will be referral from the community nutrition sites to health facilities for treatment of SAM cases. After the first 12 months of emergency response, the treatment for MAM cases will consist of lipid-based food supplementation (Plumpy’Sup) and household-level follow-up of children. A more comprehensive package will be scaled up to include preventative nutrition services. This package will be harmonized with what is being delivered in other parts of the country and is currently being refined as part of the preparation of the new health, nutrition, and population IDA operation. These activities may include (but not be limited to): (i) growth monitoring; (ii) behavior change communication around reproductive, maternal, and child health/nutrition, acute malnutrition, and chronic malnutrition; (iii) intensive counseling; (iv) food diversification education, including culinary demonstrations; (v) early stimulation activities for children ages 0 to 30 months to increase impact on early child development outcomes; and (vi) scale-up of effective prototypes currently being tested around behavior change, including messaging to mothers, more effective behavior change communication models at the community nutrition sites, training/coaching of service providers, and women’s cooperatives/income-generating modalities. ACNs will deliver these interventions at community nutrition sites and/or at household level to targeted families. There will be support to monitoring and evaluation activities, including strengthening routine monitoring systems, which including mobile data collection by the ACNs and developing a comprehensive database of beneficiaries of nutrition services that will allow for effective monitoring of delivery of services and nutritional situation of beneficiaries from community to central level. There will also be support to the ONN/U-PNNC to strengthen the grievance mechanisms to effectively collect and respond to submissions.

(c) Activity 3. Livelihood recovery grants for a total amount of MGA 180,000 for each household will be paid to the mother (who is usually the participant in the PNNC) with the objective to help recapitalize and support the households in recovering their livelihood activities. This grant will be paid in a maximum of three tranches. The first tranche of MGA 60,000 per beneficiary household will be transferred at the beginning of the program, which also corresponds to the planting season (December 2016 to February 2017) to allow households to purchase seeds and small tools and take full advantage of the first expected rains. The second and third tranche (totaling MGA 120,000, which may be transferred together, depending on the activities) will be disbursed after a period of community training designed to help households identify

14 The current national protocol for the treatment and management of MAM in an emergency context consists of daily intake of 92 g of Plumpy’Sup per child for two continuous months through four distributions, accompanied by a protective family food ration of cereals, dried pulses, and oil (when possible) as a compliance measure. In September 2016, in the South, the protective food rations consists of 400 g of cereals per person per day, 60 g of dried legumes per person per day for a household of five people on average are given to families of children with MAM during the first 15 days of distribution and the first 15 days of the third distribution.

36

the best productive opportunities for themselves. Service providers (NGOs, possibly centers for agricultural support, and so on.) will be engaged to offer, during a period of three to four months, training and orientation sessions to beneficiary households. A study of productive opportunities and seasonality will be conducted by the service provider using existing information and data, and in close consultation with local authorities and the beneficiaries themselves. Each local study will highlight options and opportunities for productive activities, as well as optimal dates for the disbursement of the grants to beneficiaries. Dates for disbursement may differ according to districts or communes. Grants will be disbursed to beneficiary households based on a simple work plan and conditioned to the realization of one preparatory activity, with support from the service provider. The disbursement will need to be approved and released by the FID. The service provider will then provide further technical support and supervision of the activities once the disbursement is made to beneficiaries.

(d) Funding for emergency response in light of an expected strong cyclone season due to La Niña in 2017. The AF will also include up to US$1.3 million to prepare for emergency response to the expected effects (floods, cyclones, and so on) of La Niña. Activities include, as previously mentioned under Subcomponent 1.3, cash-for-work activities as well as the rehabilitation of damaged small-scale social infrastructure. FID’s existing Emergency Response Manual will be applied to these activities. The implementation of the disaster response activities will be guided – as under the original project – by the project’s ESMF.

5. All activities under the AF will be implemented in the same districts, which are selected using a combined set of criteria. These criteria include: (a) vulnerability to food insecurity and prevalence of MAM and SAM in children under the age of five; (b) operational criteria, which favor areas of intervention which are contiguous and not overly scattered, as well as accessible given the precarious security situation in some of the districts; and (c) the presence of other emergency interventions by development partners and NGOs, which can in some cases lead to overlap and in other cases mutually reinforce the positive effects (for example, when linked to rural development programs). Furthermore, the AF will target communities covered by functioning nutrition sites (under the ONN) and therefore run by at least one community nutrition agent where mothers and children are already registered and attend regular sensitization and control activities. Several rounds of discussion among key partners (MPSPPW, FID, ONN, INSTAT, UNICEF, and World Bank) took place during AF preparation and a final list of districts and communes has been adopted.

6. Most beneficiary households will be selected at the start of the proposed AF. In the first year, all selected households will have children between 0 and 5 years of age (as they are active participants of the community nutrition program) and will benefit from the AF activities. In the first year, the proposed AF will target 45,000 households in the selected districts, therefore covering approximately 200,000 people all together, including 48,000 children ages five and below. The selection of beneficiary households is based on the list of mothers and children currently registered in the community nutrition centers of the targeted communities. All mothers registered at the nutrition sites (that is, having at least one child under five years of age) will be

37

automatically considered as beneficiaries of all AF activities. The ACN running the nutrition sites will also be included as a beneficiary and will therefore be offered the same benefits as the mothers. The rules for entry and exit of the program have also been agreed: (a) mothers whose children become older than five during the duration of the program will continue to benefit from the activities (even if they do not attend nutrition centers anymore); and (b) if a household leaves the program on a voluntary basis (migration, and so on), the community nutrition agent will be able to propose a new mother and child to be included for the rest of the program. Updating the list of beneficiaries will occur once every three months. In the second year, up to 20,000 additional households (about 100,000 additional beneficiaries) with children between 6 and 12 years from the same communities will also be invited to join the cash transfer program and will receive monthly cash transfers subject to compliance with co-responsibilities, where feasible.

7. Communication activities regarding all aspects of the proposed AF will be carried out in an integrated manner by the government agencies involved in the AF. A communication strategy, common to the FID, ONN, and the MPSPPW is being developed and will be agreed before effectiveness of the AF. It will aim at assisting all project processes and facilitating the implementation of the three main activities with beneficiary households, local authorities, and other partners at the local level in the selected districts. The implementing agencies envision to include community-based cultural and empowering activities through interactive theatre, radio shows, and music making.15

8. Participation of beneficiaries and social accountability will be key aspects of program implementation. Activities will be implemented with intensive participation and constant consultation of and feedback by beneficiary households. For example, it is foreseen that a grievance and redress system will be put in place (based on the experience of the FID and the system that has already been put in place) as well as a regular monitoring of the operation, including follow up of special cases (case management).

9. Collaboration between the implementation agencies. To establish the various levels of responsibilities under the proposed AF, a Memorandum of Understanding will be signed among the FID, ONN, and the MPSPPW. At the regional level, the FID will establish a new regional office in Fort Dauphin to manage the proposed AF implementation. The AF also includes some funding to strengthen the ministry’s capacity to coordinate and oversee the implementation of the program.

10. Accompanying action research and evaluation. Given the special characteristics of the South of Madagascar, its history of failed projects and programs, as well as the first time provision of regular cash transfers to beneficiaries, who are mainly women, the AF will be accompanied by an innovative anthropological and socioeconomic research program to be launched in partnership with international and local universities, as possible. In collaboration with UNICEF, this accompanying evaluation and research program has been developed between all implementing

15 A recent study on the successes and failures of projects and programs in the South of Madagascar reveals that a community-based approach is indispensable for any program that seeks to have an impact on people’s well-being. The more actively involved beneficiaries are, the more change effects can be expected. In the case of safety net programs, the cash transfers can be seen as an instrument to engage directly with households for behavior change while at the same time alleviating part of their poverty.

38

agencies during project preparation and will be partially financed by the project as well as external contributions (e.g. from UNICEF).

11. Components 2 and 3 of the SSNP will also be scaled up under the AF. US$3.1 million will be added to Component 2 of the SSNP to ensure proper management of the safety net activities under the AF. This will include the opening of an additional FID office in the South (Fort Dauphin)16 and a satellite office in Ambovombe, as well as funding for logistics, training of trainers, communication, and so on. Furthermore, Component 3 will get an additional US$1 million to enable the MPSPPW to carry out its coordination, monitoring, and evaluation functions in the South. This includes the rehabilitation of the two local offices, provision of transport, operating costs, and technical assistance to enhance the ministry’s presence in the region and facilitate coordination, particularly between the humanitarian and the development partners. It also includes funding for program monitoring and evaluation. Both components include funding for small works for the upkeep and maintenance of the project implementation agencies’ premises, which is covered under the Project’s Environmental and Social Safeguards Management Framework (ESMF).

12. Procurement arrangements will be similar to the parent project. The proposed AF will continue to be carried out in accordance with: (a) the World Bank’s ‘Guidelines: Procurement of Goods, Works, and Non-consulting Services Under IBRD Loans and IDA Credits and Grants by World Bank Borrowers’, dated January 2011 and revised in July 2014; and Guidelines on Preventing and Combating Fraud and Corruption in Projects Financed by IBRD Loans and IDA Credits and Grants, dated October 15, 2006 and revised in January 2011, and the provisions stipulated in the Legal Agreement; (b) ‘Guidelines: Selection and Employment of Consultants Under IBRD Loans and IDA Credits and Grants by World Bank Borrowers’, dated January 2011 and revised in July 2014; and (c) the provisions of the Financial Agreement.

• FID personnel from other regional offices and newly recruited personnel will staff the new regional office in Fort Dauphin. With regard to procurement readiness, the procurement officer at the head office will provide on-site procurement support during the preparation stage of the AF and continual distance support during the implementation phase. A number of procurement processes such as request and evaluation of Expression of Interest and Request for Proposals are currently ongoing to be ready for implementation as soon as effectiveness is declared.

• The procurement arrangements within MPSPPW, responsible for the implementation of Component 3, will remain the same for the AF as for the project.

• A procurement capacity assessment of the U-PNNC was carried out because it is a new implementing agency under this project. Because the initial credit of the PAUSENS will close in December 2016, only a few procurement activities remain to be followed up by the U-PNNC. The U-PNNC currently has a procurement officer and two procurement assistants, which is sufficient capacity. As such, no additional staff is required to carry out procurement activities under the proposed AF. The central

16 Fort Dauphin was selected due to its connectivity, available office space, and coordination platform with multiple partners.

39

office will initiate all procurement activities as well as update and refine the approved procurement plan by project effectiveness.

• The procurement plan has been updated and approved during negotiations.

13. FM arrangements. The three PIUs (FID, MPSPPW, and U-PNNC) will be responsible for the FM of their respective components.

• It was established that both the FID and U-PNNC have the FM staff who possess the relevant qualifications and the appropriate experience with regard to the World Bank FM procedures and requirements.

• For the ministry, the AF will necessitate the recruitment on a contractual basis of an FM specialist to support the existing team temporarily, for a period not exceeding one year.

• For the FID, the AF will necessitate the recruitment, on a contractual basis, of one FM specialist and one accountant for the regional direction of Fort Dauphin (to be created) as important SSNP activities related to the AF will be focused in this region. One accountant will be recruited at the central level to ensure the consolidation of financial data from five districts. This additional recruitment is required regarding the workload triggered by this financing and the fact that this entity will implement several safety net programs simultaneously.

• For the U-PNNC, given that this entity is currently implementing two loans of the World Bank, the recruitment of an additional accountant will be necessary to reinforce the team at the central level. No recruitment is required at the regional level as there is no significant change in targeted zones. The conclusion of the FM assessment is that the proposed FM arrangements meet the World Bank’s minimum requirements under OP/BP 10.00 Investment Project Financing. The overall residual risk rating is Substantial.

14. Safeguard arrangements. No change

40

Annex III: Geographic Targeting of the Additional Targeting

Map of Southern Districts with Malnutrition Rates

Source: Foibe Taon-tsarin-tany Malagasy - FTM, ONN, UNICEF(Geographical Institute of Madagascar)

41

Map of Geographical Targeting of Additional Financing

Source: Foibe Taon-tsarin-tany Malagasy - FTM, ONN, UNICEF(Geographical Institute of Madagascar)

42

Annex IV: Revised Estimate of Project Costs (US$, millions) Component Current

Amount AF Revised Total

Allocation Component 1. Building a Safety Net for the Poor in

Selected Rural Areas

Subomponent 1.1: Establishing a Productive Safety Net Program (FID)

14.7 0.0 14.7

Subcomponent 1.2: Expanding the Human Development Cash Transfer Pilot (FID)

11.5 0.0 11.5

Subcomponent 1.3: Early Recovery Response to Natural Disasters Of which

1.3.1 Safety nets (FID) 1.3.2 Nutrition services (PNNC) 1.3.3 Livelihood grants (FID) 1.3.4 Disaster response (FID)

4.0 30.9

Of which 17.1

8.5 4.0 1.3

34.9

Component 2: Strengthening Safety Net Administration, Monitoring, and Social Accountability (FID)

6.5 3.1 9.6

Component 3: Building the Institutional Capacity for Coordination, Monitoring and Evaluation of the Social Protection System (MPSPPW)

3.3 1.0 4.3

TOTAL 40.0 35.0 75.0

Table 4.1. Simulated impacts of Cash Transfer Program in the South

Benchmark Scenario 1 Scenario 2 Scenario 3 UCT (benefit size in Malagasy ariary per month) 30,000 30,000 30,000 20,000 CCT (benefit size in Malagasy ariary per month) 20,000 20,000 30,000 20,000 UCT (number of months) 12 18 12 12 CCT (number of months) 24 18 17 24 Livelihood recovery (onetime grant) 180,000 180,000 180,000 180,000

UCT (number of households covered) 45,000 45,000 45,000 45,000 CCT (additional number of households covered 20,000 19,200 16,200 31,250 Total cost (MGA, thousands) 69,375,000 69,390,000 69,390,000 69,375,000 Total cost (US$) 23,125,000 23,130,000 23,130,000 23,125,000 Impact on Consumption of Beneficiaries Increase in annual consumption (%) 67 72 86 57 Size relative to food poverty gap (%) 16 17 16 13 Coverage of Beneficiaries % of food-poor households in three regions 32 31 30 37

Note: UCT = Unconditional Cash Transfer.

43

Annex V: Summary of the Human Development Cash Transfer Program (HDCT) (Component 1.2 of the Social Safety Net Project)

1. Objective of the HDCT. The HDCT aims to provide income support to the poorest families to smoothen their consumption levels and to enhance their children’s human capital by requiring children ages 6 to 12 years to attend primary school regularly and by enhancing families’ health, nutrition, and early childhood practices. The project beneficiaries will be extremely poor households (39,000 households) living in communities in selected regions: Atsinanana (East), Atsimo Andrefana (South), Haute Matsiatra (Central), Vatovavy Fitovinany (Southeast), and Vakinankaratra (Central).

2. The program cycle is composed of the following process steps:

(a) Targeting. A combination of geographic, community, and household targeting

(b) Household data processing. The household data are entered into a computerized system. These beneficiary data are then transferred to the FID where they are used to manage the program’s implementation. The FID is then responsible for updating the beneficiary data (in accordance with co-responsibility compliance data, household changes (birth/death of a household member, households moving, and so on).

(c) The enrollment is mainly to determine which household member will collect the transfer (usually the female head of the households) as well as an alternative recipient if the original person is unavailable and to collect/update the household information required for the payment agency. The process is also to train and inform the household about the rules, rights, and obligations of the HDCT Program and about how to report updates, make complaints, and communicate with the program management and implementing agencies.

(d) Payments. Transfers are paid every two months through financial intermediary institutions. The level of cash transfers depend on the household composition and compliance with co-responsibilities. A basic monthly benefit of MGA 10,000 to support all beneficiary households in meeting their basic needs is provided. An education-linked monthly benefit of MGA 5,000 per child attending primary school is provided to a maximum of two children attending primary school. A onetime school enrollment benefit of MGA 20,000 is provided to each newly enrolled beneficiary household at the beginning of a school year to cover the expenses of enrolling their children in school.

(e) Compliance monitoring and complaints and updates. Communication about the program continues throughout the program cycle in different styles and with different audiences. Community participation is ensured through social protection committees established at the fokontany level, as well as mother leaders, who are elected by the beneficiaries to provide a direct link between the beneficiaries and the program implementers.

44

3. Mother leaders. The beneficiary households will elect a lead mother to be the link between the beneficiaries and the program. This lead mother will be responsible for three things in principle. First, they have an operational role by keeping beneficiaries informed about the program’s rules, responsibilities, and procedures including payment schedules and complaint mechanisms. Second, they are responsible for organizing bimonthly meetings to learn about and discuss issues related to nutrition, health, essential family practices (Pratiques Familiales Essentielles), early childhood development (for children ages 3 to 6 years), and other topics of their interest. These meetings will also serve as a place to share opinions and information among community members in the same environment who face the same challenges. Third, the lead mothers will play a role during the paydays in helping and following up on the ‘nudges’ (behavioral interventions) that are being developed. For all this, the social organizers and UNICEF (on essential family practices) will train the mother leaders and others based on a training plan developed by the FID.

4. Collaboration among various government and nongovernment agencies. The implementation of the HDCT requires the interaction of the MPSPPW and FID with the MEN and ONN. The MEN will provide school attendance reports and the ONN will strengthen beneficiaries’ nutritional awareness. The MPSPPW will monitor the program’s activities, while UNICEF will support the implementation of the essential family practices.

5. Communications and outreach. The HDCT will include a communications campaign designed to support the rollout of the HDCT along its program cycle. The main objectives of the communications campaign are to:

(a) communicate the objective of the program widely;

(b) disseminate information about the targeting process and criteria; and

(c) ensure that, during the enrollment process, the beneficiaries are fully informed about their rights, obligations, (co-)responsibilities, cash transfers, and the functioning of the complaint process, and that the information is communicated regularly and several times to ensure full understanding and engagement over time.

45

Annex VI: Summary of the National Community Nutrition Program (PNNC) Context

1. The Government’s flagship PNNC started as a small effort in the 1990s to protect the population from the negative impacts of periods of food insecurity. Today, through a network of about 7,000 rural sites, the PNNC now reaches 2.4 million poor mothers and children under five years of age. The program is an operational unit of the country’s ONN, charged with the overall mandate to coordinate the multi-sectoral nutrition agenda. The results from the multi-round impact evaluation (1999, 2003, 2007, and 2011) showed that the PNNC had a significant impact when it comes to improving key outcomes related to acute malnutrition, for the target population. Since 2012, recognizing the high levels of stunting in the country, the PNNC has expanded its scope to start delivering interventions to reduce chronic malnutrition. The sites, run by nutrition agents, chosen by their communities, also serve as critical links/delivery platforms for a number of different programs at the village level including for health, agriculture, education, social protection, and water and sanitation. The program addresses a number of different issues over time and has proven resilient and responsive to the country’s multiple crises. The World Bank currently supports the model in nine regions of the country through a multi-sectoral IDA operation, the PAUSENS Project (P131945).

Overview of Design

2. The core element of the PNNC is a network of community nutrition sites established in land granted by the community and run by ACNs chosen by their respective communities. Local NGOs contracted by the National Nutrition Program (at the central level) are responsible for compiling data from the ACN, writing technical and fiduciary reports that highlight process and key challenges/issues, and paying the ACN (who receive a small monthly amount for their efforts on running the sites). The reports are compiled and submitted on a monthly basis to the regional nutrition offices (part of the national nutrition program), which verify the submission and pay the NGOs based on the completeness of those submissions. Each NGO is responsible for managing about 10–12 sites. In addition, every four months, ACNs are given the opportunity to earn a ‘performance bonus. The regional nutrition offices go to the communities and measure the performance based on a scorecard of indicators that includes interviews with mothers.

Results

3. The results of a long-term impact evaluation spanning 1998–2011, (Fernald, Galasso, and Weber 2016) showed that the program has had a significant impact on weight-for-age (underweight) for young children with protective effects during bouts of fragility. While under the model, the ACNs receive a very small payment linked to output results every month, even when the state has not been able to pay due to various crises, the majority of the sites have remained operational, pointing to resilience in the system. With the current crisis in the South due to the severe drought, the existing community nutrition sites are already in use with a number of programs rapidly scaling up the screening and treatment of acute malnutrition. This includes follow-up of children at the household level by agents and ensuring a referral link from the community to the health center for treatment of severe cases.

46

Addressing Stunting and Improving Early Child Development through PNNC

4. In 2012, the long-term evaluation results and the emerging global evidence on stunting, spurred the Government to drastically rethink its strategy and scope on multiple fronts with the support of the World Bank and other donors. First, the National Strategy was refocused to prioritize the first 1,000 days with an emphasis on scaling up prevention efforts at the community level (through the National Nutrition Program) and the package of interventions delivered at the community nutrition sites was refined. Second, the program got a deeper understanding of behavior change through counseling. In this context, the learning from their South-South exchange with Building Resources Across Communities, Bangladesh has been adapted, and additional community nutrition workers are providing intensive counseling at the household level. Third, the program is testing the cost-effectiveness of lipid-based nutrient supplements for pregnant women and children as a preventive tool. Fourth, to address multiple pathways to improve child development outcomes, early stimulation activities (for children ages 6 to 30 months) have been integrated at the household level by teaching caregivers to effectively interact with children. Using IDA resources, the Government is piloting and testing the integrated delivery of these interventions through a randomized control trial with the intention of scaling up effective activities through the existing PNNC (initial results of the pilot will be available in November 2016). Using innovation processes, specifically the human-centered design approach, the Government is also unpacking the ‘black box’ around beneficiary barriers and aspirations influencing behavioral change and rethinking how frontline workers are trained, monitored, and motivated. This evidence and learning will contribute to enhancing the community nutrition package in Madagascar.

47

Annex VII: Key Results from the Health and Education Public Expenditure Review 1. The distributional analysis of the health and education sector expenditures shows that the poorest regions have the lowest spending for both sectors. Moreover, considering that two-thirds of the population live in rural areas, the low amount of education and health expenditure in these areas is indicative of large inequities in the distribution of public resources, largely driven by unequitable distribution of human resources across the country. During the crisis of 2009–2013, cuts in public spending to social sectors resulted in large increases in household spending, particularly on education. This has had far-reaching consequences with regard to enrollment in primary schools, learning outcomes, and utilization of health services. Better equity in the distribution of spending will require that trained and qualified teachers are equally distributed across the country and that additional health human resources target the neediest areas. Finally, severe data limitations due to the lack of a recent census and more recent household data at the district level prevent a finer analysis of the distribution of spending by poverty level. The distribution of government current expenditures by region indicates some regression, with some leveling out during the most recent period. Expenditure shares by type of residence have evolved in favor of the large urban centers (+14 percent), mostly at the expense of other urban areas and peri-urban rural areas. Rural areas (including the peri-urban areas) received less than one-third of the expenditures.

2. Overall, the MoH expenditures are strongly negatively correlated to poverty rates and strongly positively correlated with per capita income. Even when restricting expenditures to primary health care, regional distribution is strongly regressive. Benefit Incidence Analysis on non-wage expenditures executed by the MoH reveals that benefits are found to be regressive with the average benefits going to individuals in the richest quintile, two to four times higher than those going to individuals in the poorest two quintiles. Benefit Incidence Analysis performed on wage expenditures on personnel in health facilities show that the richest quintile benefits 3.6 times more than the poorest quintile and at least twice as much as households in any other quintile. Analysis of expenditure shares by type of residence shows that less than 5 percent goes to rural communes. On average 36 percent goes to large urban centers, 46 percent to smaller urban areas, and 13 percent to the semirural or peri-urban areas.

Figure 7.1: Estimated Distribution of Benefits from MoH Non-wage Expenditures

8.2 8.9 13.5 13.2

29.6

0.05.0

10.015.020.025.030.0

Lowestquintile

Secondquintile

Thirdquintile

Fouthquintile

Richest

MG

A, B

illio

n 20

13

Estimated Distribution of Benefits from MoH Non-wage Expenditures

48

Figure 7.2: MoH Recurrent Expenditure Per Capita Ranked by 2005 Poverty Rates

Sources. Data from Ministry of Finance and Budget, INSTAT (population), EPM (poverty).

3. Households are financing an increasing share of the total costs of education. Since the crisis, a greater number of households with a child in Grade 2 have had to pay enrollment fees, parent-teacher association contributions, and monthly school fees. In addition to being more frequent, the amount paid in school fees have generally increased by 26 percent on average. This has had direct repercussions on enrollment. Indeed, financial problems are the first reason given by households to explain dropout. Over 2006–2008, current public expenditure contributed, on average, to 73 percent of expenses of one child enrolled at school, whereas household expenditure represented, on average, 27 percent. However, from 2009 to 2013, the share of current public expenditure was, on average, 59 percent of total spending per child enrolled. This reflects a substantial increase in the share of the costs per child enrolled financed by households. The increase was higher in Atsimo Andrefana, Menabe, and Vatovavy Fitovinany, three regions that are particularly vulnerable.

Figure 7.3: Financing of Primary Education Share of Household and Public Financing

0102030405060708090100

0.02.04.06.08.0

10.012.0

AN

ALA

M…

BO

ENI

DIA

NA

ALA

OTR

…M

ENA

BE

MEL

AK

YB

ON

GO

L…IT

ASY

BET

SIB

O…

HA

UTE

…SA

VA

AN

OSY

VA

KIN

A…

ATS

IMO

-…A

MO

RO

N…

IHO

RO

MB

EA

TSIN

AN

…A

NA

LAN

…SO

FIA

VA

TOV

A…

AN

DR

OY

ATS

IMO

-…

%

Tho

usan

d of

201

3 A

riar

y

Average 2006-2008Average 2009-2013Poverty Ratio 2005Poverty Ratio 2010

0%10%20%30%40%50%60%70%80%90%

100%

2005 2010

Avg. current public exp. per child Avg. household exp. per child

49

Figure 7.4. Current Public Expenditure per Capita in Education by Region Ranked by 2005 Poverty Rates

Source: Data from Ministry of Finance and Budget, INSTAT (population), EPM (poverty).

0

20

40

60

80

100

0.0

10.0

20.0

30.0

40.0

50.0 %

Tho

usan

d of

201

3 A

riar

y

Average 2006-2008

Average 2009-2013

Poverty Ratio 2005

Poverty Ratio 2010